Prostate Cancer Local Staging with Magnetic Resonance Imaging

Abstract

Accurate determination of the local stage of prostate cancer is crucial for treatment planning and prognosis. The primary objective of local staging is to distinguish between organ-confined and locally advanced disease, with the latter carrying a worse clinical prognosis. The presence of locally advanced disease features of prostate cancer, such as extra-prostatic extension, seminal vesicle invasion, and positive surgical margin, can impact the choice of treatment. Over the past decade, multiparametric MRI (mpMRI) has become the preferred imaging modality for local staging of prostate cancer and has been shown to provide accurate information on the location and extent of disease. It has demonstrated superior performance compared to staging based on traditional clinical nomograms. Despite being a relatively new technique, mpMRI has garnered considerable attention and ongoing investigations. Therefore, in this review, we will discuss the current use of mpMRI on prostate cancer local staging.

Introduction

Worldwide, prostate cancer is the most commonly diagnosed malignancy and the fifth leading cause of cancer death among men¹. In the United States, it is the second leading cause of cancer deaths in men, in which African American men suffer from the highest incidence and mortality rates². With prostate cancer affecting such a large majority of the male population, assessing the presence and extent of clinically significant prostate cancer and characterizing the risk of future progression is a priority. Over- or under-treatment can be avoided by accurately distinguishing low- vs. high-risk cases and tailoring treatment accordingly.

Magnetic resonance imaging (MRI) has become increasingly used in detecting clinically significant prostate cancers and can facilitate the triage of patients for biopsy³. Additionally, MRI provides valuable information for staging tumor extent and monitoring treatment response⁴. Prostate MRI can aid in staging evaluation by verifying organ-confined status and establishing the location and extent of the tumor in patients being considered for radiation therapy or surgery⁵. It is often used for surgical planning, especially for defining presence of tumor in the neurovascular bundle and determining the need for nerve removal. While the practice of verifying organ-confined disease status with prostate MRI prior to nerve-sparing radical prostatectomy is well-supported by evidence^{6, 7}, it has yet to be fully integrated into major practice guidelines. According to the American College of Radiology (ACR) appropriateness criteria, prostate MRI is recommended for all intermediate- and high-risk patients to assist in treatment planning⁸. Meanwhile, the updated guidelines from the European Association of Urology/European Association of Nuclear Medicine/European Society for Radiotherapy and Oncology/European Society of Urogenital Radiology/International Society of Geriatric Oncology (EAU-EANM-ESTRO-ESUR-SIOG) suggest prebiopsy prostate MRI for local staging for any risk group⁹. Given there is currently no consensus on how MRI should be deployed in clinical practice, in this review, we will describe the current data in staging of local prostate cancer using MRI, specifically in the evaluation of extra-prostatic extension (EPE), seminal vesicle invasion (SVI), and positive surgical margin (PSM).

Management of Locally Advanced Prostate Cancer

Patients with locally advanced prostate cancer and with high-risk features are advised to undergo definitive local therapy. Treatments for these patients typically includes external beam radiotherapy combined with long-term androgen deprivation therapy (ADT)^{10, 11}. External beam radiotherapy combined with ADT provides patients with better survival outcomes compared to ADT alone and external beam radiotherapy alone¹¹. Androgen deprivation therapy can be administered neoadjuvant, concurrently, or adjuvant when combined with radiation therapy¹². For a subset of patients (such as being clinically node positive, having 2 out of 3 of the following criteria: clinical stage T3 or T4, PSA level of ≥40 ng/mL, or a Gleason score of ≥8), radiation and androgen deprivation therapy, along with a two-year course of concurrent abiraterone acetate and prednisone may be recommended as well^{12, 13}. However, there are toxicities associated with long term continuous ADT, such as hot flashes, insomnia, and decreased libido¹⁴. For individuals at a very high risk, particularly those who are younger, radical prostatectomy in conjunction with extended pelvic lymph node dissection is another treatment option due to favorable cancer survival rates¹⁰. Studies indicate 5-year prostate-specific antigen (PSA) relapse–free survival rates ranging from 55%–71% and 10-year prostate cancer–specific survival rates from 72%–92%^15-18. However, the most effective treatment for these patients is still currently being actively studied, and previous meta-analyses have reported inconsistent results regarding survival rates for different treatment options^{19, 20}. Nonetheless, it is important to note that the identification of EPE or SVI on MRI may lead to a discussion regarding the impact of T3 disease on pathology, the potential requirement for adjuvant therapies, and the possible consideration of nerve sparing of the ipsilateral bundle. Accurate local staging is, therefore, essential in deciding the optimal treatment strategy and prognosing patient outcomes.

The Prostate Imaging Reporting and Data System (PI-RADS)

The implementation of the Prostate Imaging Reporting and Data System (PI-RADS) has significantly enhanced the consistency and standardization of prostate cancer reporting²¹. PI-RADS is used to characterize and assess all focal intraprostatic lesions seen on multiparametric MRI (mpMRI). It offers a set of criteria to evaluate the likelihood that a lesion detected through mpMRI indicates clinically significant cancer. The guideline also includes technical standards for scanning hardware and protocols for image acquisition. To obtain the best images and results, PI-RADS guideline recommended acquiring images using a 3 T over 1.5 T scanners. The use of an endorectal coil is optional but may be considered for better image quality. In addition to T2WI, obtained in the straight or oblique axial plane and at least one additional orthogonal plane, high-b-value DWI and DCE MRI should be routinely included in image acquisition protocols.

PI-RADS classifies prostate lesions on a five-point scale based on the probability of clinically significant cancer (Figure 1). The primary sequence recommended for determining the category of the peripheral zone lesions is DWI (Figure 2), while for transition zone lesions, T2WI is used (Figure 3). In general, clinically significant cancer can be defined as a lesion that is predicted to have an International Society of Urological Pathology (ISUP) Grade Group of 2 or higher, with either a volume ≥ 0.5 mL or extra-prostatic extension. Studies have found the cancer detection rates for PI-RADSv2 may be variable and category 3, 4, and 5 assessments in detecting clinically significant prostate cancer range from 12-23%, 39-60%, and 72-83%, respectively^{3, 22, 23}. The differences in cancer detection rates may be due to intra- and inter-reader variability related to reader experiences and subjectivity^24-26. However, some studies have shown the negative predictive value of PI-RADSv1 and v2 in detecting clinically significant cancer range from 86-91%^{27, 28}. Therefore, mpMRI of the prostate is generally an accurate test for ruling out clinically significant prostate cancer. PI-RADSv2.1 was refined and introduced in 2019. One meta-analysis with 11 studies showed the clinically significant cancer detection rates for PI-RADSv2.1 category 3, 4, and 5 assessment range from 13-27%, 43-61%, and 76-97%, respectively²⁹. Its true detection performance is still under investigation, and large prospective studies are needed³⁰. Radiologists should be aware of the inherent limitations of PI-RADS guidelines, specifically regarding the inconsistency in the published literature on the importance of a PI-RADS 3 lesion and the necessity of biopsies for all PI-RADS 3 lesions^{3, 31}.

Figure 1.

PI-RADS assessment categories

Figure 2:

Multiparametric MRI of a 74-year-old man with a serum prostate specific antigen level of 8.1 ng/mL. The lesion (arrows) was 1.3 cm in the midline apical-mid peripheral zone and assigned PI-RADS category 4. Targeted biopsy revealed Gleason score 7 (4+3) prostate adenocarcinoma. T2-weighted imaging (A), apparent diffusion coefficient map (B), high b-value (b=1500 s/mm²) diffusion-weighted imaging (C), and dynamic contrast enhanced imaging (D).

Figure 3:

Multiparametric MRI of a 62-year-old man with a serum prostate specific antigen level of 14.7 ng/mL. The lesion (arrows) was 1.5 cm in the right base anterior transition zone and assigned PI-RADS category 5. Targeted biopsy revealed Gleason score 7 (3+4) prostate adenocarcinoma. T2-weighted imaging (A), apparent diffusion coefficient map (B), high b-value (b=1500 s/mm²) diffusion-weighted imaging (C), and dynamic contrast enhanced imaging (D).

Evaluation of Extra-prostatic Extension

For patients with prostate cancer, it is crucial to undergo a comprehensive risk assessment and staging process to ensure that they receive the best possible counseling and treatment. The American Joint Committee of Cancer (AJCC) tumor-node-metastasis (TNM) system, last revised in 2018, is the most commonly employed staging system for prostate cancer³² (Table 1). Prostate cancer with EPE are staged as T3a according to the TNM staging system. The presence of EPE is associated with a higher risk of PSM, biochemical recurrence, metastatic disease, and lower survival rate after undergoing radical prostatectomy^{33, 34}.

Table 1.

American Joint Committee on Cancer (AJCC) tumor-node-metastasis (TNM) Staging Manual

Primary Tumor (T)
Clinical T (cT)	cT Criteria
T0	No evidence of primary tumor
T1	Clinically inapparent tumor not palpable
T1a	Incidental tumor in < 5% of tissue resected
T1b	Incidental tumor in > 5% of tissue resected
T1c	Tumor identified by needle biopsy found in one or both sides, but not palpable
T2	Tumor is palpable and confined within prostate
T2a	Tumor involves one-half of one side or less
T2b	Tumor involves more than one-half of one side but not both sides
T2c	Tumor involves both sides
T3	Extra-prostatic tumor that is not fixed or does not invade adjacent structures
T3a	Extra-prostatic extension (unilateral or bilateral)
T3b	Tumor invades seminal vesicle(s)
T4	Tumor invades external sphincter, rectum, bladder, levator muscles, and/or pelvic side wall
Tx	Primary tumor cannot be assessed
Pathologic T (pT)	pT Criteria
T2	Organ-confined
T3	Extra-prostatic extension
T3a	Extra-prostatic extension (unilateral or bilateral) or microscopic invasion of bladder neck
T3b	Tumor invades seminal vesicle(s)
T4	Tumor invades external sphincter, rectum, bladder, levator muscles, and/or pelvic side wall
Lymph Nodes (N)	N Criteria
N0	No positive regional lymph nodes
N1	Regional node metastases – including pelvic, hypogastric, obturator, iliac, and/or sacral
Nx	Regional nodes were not assessed
Distant Metastases (M)	M Criteria
M0	No distant metastasis
M1	Distant metastasis
M1a	non-regional lymph nodes (outside true pelvis)
M1b	Bone(s)
M1c	Other site(s) with or without bone disease

On MRI scans, the prostatic capsule is mainly evaluated on axial T2W MRI since it has a higher spatial resolution compared to DWI and DCE images. The capsule appears as a hypointense rim surrounding the prostate gland (Figure 4). It is best visible in the middle 70-80% of the prostate gland craniocaudally, whereas at the apex and base (i.e., the inferior and superior-most portions of the gland), it is often difficult to visualize. EPE can generally be defined as bulging of the capsule, irregular or ill-defined capsular surface, thickened neurovascular bundle, or visible invasion of the surrounding structures (i.e., bladder neck and rectal wall)³⁵ (Figure 5). Pesapane et al.³⁶ suggested that MRI staging criteria of EPE can be based on the concept of a linear, stepwise progression of cancer growth.. This concept allowed the researchers to define certain characteristics of EPE into stages of “early” and “late”. The “early” stages include: : disruption of the capsule, bulging of the contour that surrounds the prostate, and then eventually an irregular prostate margin. In the early stages, as the cancerous cells spread into the capsule, they disrupt the normal signal typically seen on the MRI. As tumor cells continue to grow beyond the capsule and beyond the smooth muscle of the prostate, it reflects on the MRI as an extension or ‘bulging’ of the margin of the prostate contour. . The “late” stages seen on T2WI include: a rise and fall of the contour surrounding the prostate , obliteration of fat surrounding the prostate , and a obliteration of the angle between the rectum and the prostate. As the cancerous cells continue to grow, they extend past the prostate capsule, into the surrounding fat, and into the rectoprostatic angle. In the final stages of EPE, an overt macroscopic periprostatic mass is present.

Figure 4:

Axial T2W MRI demonstrating mid prostate gland with prostatic capsule and neurovascular bundle. The prostatic capsule (arrowheads) appears as a hypointense rim surrounding the prostate gland. The neurovascular bundle (arrow) is a tubular structure located posterolateral to the prostatic capsule and adjacent to the peripheral zone. It mediates erectile function and continence. It is an expansion of the pelvic hypogastric plexus that contains the cavernous nerves of the penis and prostatic branches of the inferior vesical artery and prostatic veins.

Figure 5:

MRI of a 67-year-old man with a serum prostate specific antigen level of 72.7 ng/mL. PI-RADS category 5 lesion (asterisk) in the left apical-mid peripheral zone with frank extraprostatic extension (arrows) and possible rectal wall involvement (arrowheads). T2-weighted imaging (A), apparent diffusion coefficient map (B), and high b-value (b=1500 s/mm²) diffusion-weighted imaging (C).

At histopathology, EPE in prostate cancer denotes the spread of tumor cells beyond the fibromuscular pseudocapsule of the prostate gland and into the adjacent periprostatic soft tissues, such as the periprostatic fat³⁷.

A common criteria used to determine the extent of EPE was introduced by Epstein et al³⁸who characterized the extension as focal or established. Focal EPE extension was classified as a few cancerous glands outside of the protstate. Whereas established (or extensive) EPE was classified as cancerous cells present in more than a few glands. Currently, the criteria for identifying focal/minimal vs nonfocal/established/extensive EPE invasion is subjective to the pathologist. However, many still use the criteria established by Epstein.

This clarification in focal versus extensive EPE is important because extensive EPE is believed to be more frequently associated with PSM compared to focal EPE³⁹. The most understood mechanism of how these tumors cells extend from the parenchyma to the soft tissue is via perineural invasion. This mechanism is consistent with the fact that EPE is most often found at the posterolateral areas of the prostate⁴⁰. There is very little capsule present in the base and apex of the prostate, so at these anatomical locations EPE is diagnosed when the cancer extended into the periprostatic fatty tissue³⁵.

In cases where patients opt for surgery, the preoperative assessment of the location of pathologic EPE is crucial in determining the surgical approach to be taken. A wider excision with removal of neurovascular bundles may be attempted to increase the likelihood of a negative margin⁴¹. However, while this more aggressive approach improves cancer control, it results in higher rates of urinary incontinence and erectile dysfunction⁴². It is important to closely examine the apex of the prostate. If cancer affects the external urethral sphincter, there is a risk during surgery of damaging the sphincter, which could lead to urinary continence. Additionally, the presence of a tumor in this area may impact the use of radiation therapy. Therefore, it is essential to have accurate assessment of pathologic EPE before treatment to optimize clinical decision-making and minimize the risk of side effects. Because EPE is difficult to determine intraoperatively, the diagnosis must be made preoperatively. This risk is traditionally assessed using nomograms, such as the Partin tables⁴³ or the Memorial Sloan Kettering Cancer Center [MSKCC] nomogram⁴⁴, which consider clinical variables such as PSA levels, digital rectal exam, and Gleason score at biopsy. However, these traditional methods do not incorporate imaging modalities and have been found to be less effective than MRI in identifying the location and extent of EPE^{45, 46}. As a result, MRI can potentially improve the accuracy of EPE localization and help physicians and patients make more informed decisions regarding treatment options.

Standardized MRI-Based EPE Evaluation

The PI-RADS guideline provide a brief overview of what radiologists should evaluate on mpMRI for EPE. The guideline suggests assessing features such as asymmetry or invasion of the neurovascular bundles, bulging prostatic contour, irregular or spiculated margin, obliteration of the rectoprostatic angle, tumor-capsule interface > 10 mm, breach of the capsule with evidence of direct tumor extension or bladder wall invasion²¹. However, there is little guidance on the utility or importance of these findings. Studies have evaluated the performance of the PI-RADS for detecting EPE and have found moderate sensitivity (ranging from 40% to 88%) and specificity (ranging from 75% to 83%)^47-50. One meta-analysis with 45 studies demonstrated a sensitivity of 61% and a specificity of 88% for MRI in detecting EPE ⁵¹. In other words, the current data suggest that MRI has a high specificity but low sensitivity for EPE evaluation. Of note, one study found that PI-RADS categories 3 or less assessment could confidently rule out the presence of EPE irrespective of clinical risk group, with an overall sensitivity of 99% and negative predictive value of 98%⁵². The variations in the detection metrics are likely due to the lack of standardized EPE criteria from the PI-RADS guideline and the difference in readers experiences, with one study showing moderate inter-reader agreement (κ = 0.45)⁴⁷.

To address some of the shortcoming of PI-RADS on predicting EPE, few groups have reported use of systematic EPE evaluation methods^{53, 54}. One of these systems was developed at the National Cancer Institute (NCI) where it has been actively used for over a decade and half⁵⁵. The NCI EPE grading system categorizes curvilinear contact length ≥ 1.5 cm or capsular bulge and irregularity seen on MRI as grade 1, both features together as grade 2, and frank capsular breach as grade 3 (Figure 6). In the prospective study with 553 participants, a higher NCI EPE grade was associated with an increased risk of EPE at pathology. Grade 1, grade 2, and grade 3 had positive predictive value of 24%, 36%, and 66% for detection of EPE on histopathology, respectively. Furthermore, clinical features combined with the grading system predicted pathologic EPE better than imaging alone, but the study did not specify how the clinical features can be incorporated into the scoring system.

Figure 6.

Axial T2WIs with MRI-derived extra-prostatic extension (EPE) grade for prediction of pathologic EPE. EPE grade 1 defined as either a curvilinear contact length ≥ 1.5 cm (arrows, A), or a capsular bulge or irregularity (arrows, B). EPE grade 2 is defined as curvilinear contact length ≥ 1.5 cm plus capsular bulge or irregularity (arrows, C). EPE grade 3 is defined as a well-defined breach of the prostate capsule with tumor extension into periprostatic space or invasion of adjacent anatomic structures (arrows, D).

Various studies have validated and compared the effectiveness of different MRI-based systems in assessing EPE, most commonly the NCI EPE grade, European Society of Urogenital Radiology (ESUR) score, Likert scale, and tumor capsule contact length^56-58. These studies have found that all these MRI-based criteria demonstrated moderate diagnostic performance for EPE detection at pathology. Specifically, Park et al. evaluated all four MRI-based EPE criterias⁵⁷. The area under the receiver operating characteristic curve (AUC, 0.77-0.85) and intra- and inter-reader agreement (κ = 0.61-0.74) for all of these systems were similar. The NCI EPE grade, however, demonstrated the highest correlation with histopathology (Spearman correlation coefficient of 0.42-0.55). The NCI EPE grade incorporates more objective parameters than the ESUR score and is further enhanced by including the measurement of tumor capsule contact length, which serves as a quantitative marker. As a result, the EPE grade is a more objective approach compared to the ESUR score and the Likert scale. Another study also showed the reader with less prostate imaging experience had lower sensitivity using the Likert scale and ESUR score compared to the NCI EPE grade, suggesting that the latter method is less reader experience dependent⁵⁶. However, this is only reported in one study, and EPE grade does include subjective criteria that may be affected by reader’s experience (i.e., capsular bulging and irregularity). One study by Reisaeter et al. found the inter-reader agreement for the NCI EPE grade and Likert scale was fair with weighted κ of 0.47 and 0.45, respectively⁵⁸. Overall, the NCI EPE grade seems to perform as well as other systems but with the advantage of standardization. It is evident that combining clinical risk factors with imaging characteristics is necessary to improve the accuracy of predicting EPE and making informed treatment decisions⁴⁵. Gandaglia et al. proposed a model to predict EPE using clinical variables with mpMRI and systematic biopsy results⁵⁹. The study suggested the inclusion of mpMRI data may improve the discrimination of clinical models for EPE modestly (70%, 95% confidence interval [CI]: 65-74% vs. 67%, 95% CI: 61-70%). An external validation of the model showed no significant performance improvement compared to the MSKCC model (72% vs. 70%, p = 0.3), but a significant improvement compared to the Partin tables (72% vs. 61%, p < 0.001)⁶⁰. Updated models that incorporate mpMRI information with greater generalizability and performance still need to be developed and investigated.

The experience and expertise of the interpreting radiologist are also essential in the detection of EPE. Radiologists who are experienced in interpreting prostate MRI images are more likely to detect EPE accurately than those with less experience. Therefore, it is essential to have radiologists with specialized training and experience in interpreting prostate MRI scans to improve the detection of EPE^{61, 62}. Incorporating a standardized EPE reporting system such as the NCI EPE grade into PI-RADS would be a helpful first step to further improve the uniformity of mpMRI interpretations, and to transition from a overly simplistic binary assessment (i.e., EPE present vs. absent) to a probability risk assessment. If the grading system outcomes were placed as a footnote within reports, the grade would be readily associated with its corresponding quantitative risk of EPE. It is probable that enhancements in the consistency between readers can be achieved with the incorporation of artificial intelligence-based solutions that offer quantitative evaluations of EPE risk as well. Additionally, it is important for radiologists to recognize the limitations of mpMRI in predicting EPE. For example, even in cases of visible gross extension on mpMRI using the NCI EPE grade, only 66% were found to have EPE on pathological examination, indicating potential false-positive results. Inflammation, desmoplastic reaction, or changes related to biopsy-induced trauma can contribute to such false positives. Moreover, the microscopic nature of EPE as a histopathological definition makes it challenging to accurately predict on mpMRI. However, positive predictive value may be influenced by disease prevalence and positive predictive value of mpMRI in detecting EPE could be as high as 89% in high-risk patients⁶³. Nonetheless, it is crucial to approach the diagnosis of EPE with caution when using mpMRI.

Evaluation of Seminal Vesicle Invasion

SVI is another important factor for local cancer staging and prognosis due to an increased risk of lymph node metastasis⁶⁴. SVI stages prostate cancer as T3b of the TNM staging system; tumors with SVI are classified as locally advanced and have very high risk for progression or recurrence⁶⁵. Prostate cancer patients with SVI were found to have a 32% 7-year survival rate, while men without SVI had a 67% 7-year survival rate⁶⁶.

According to the ISUP Consensus, SVI is defined as infiltration of tumor cells that did not originate from the seminal vesicles or ductus deferens present in the muscular layer on histopathology⁶⁷. A study conducted by Ohori et al categorized the mechanism of prostate cancer invasion into the seminal vesicle into three categories. Type 1 SVI was defined as a direct mechanism in which the prostate cancer spread along the ejaculatory duct and into the seminal vesicle. Type 2 SVI was characterized as the spread of the cancer through the prostate capsule and into the seminal vesicle. Researchers further split type 2 into two subgroups. Type 2A invasion occurs between the base of the prostate and the seminal vesicle, while type 2B invasion occurs when the prostate cancer grows from the periprostatic nerve into the seminal vesicle. “The most uncommon mechanism of spread is Type 3 SVI in which there are deposits of cancer in the seminal vesicle with no contiguous primary cancer in the prostate.” ⁶⁸. SVI can be identified by certain characteristics on MRI. These may include hypointense filling defects within SVs on T2W MRI with early contrast enhancement on DCE MRI and restricted diffusion (hypointense on ADC maps, hyperintense on high b DW MRI) within and/or along the seminal vesicle. Additionally, the obliteration of the angle between the base of the prostate and the seminal vesicle, as well as direct tumor extension from the base of the prostate into and around the seminal vesicle are among the MRI features of SVI⁵³ (Figure 7). The PI-RADS guideline recommended that high spatial resolution T2WI is essential for assessment of SVI²¹. Additionally, some studies have suggested that abstaining from ejaculation 3 or more days, especially in men older than 60 years, before undergoing an MRI examination may enhance seminal vesicle assessment. Abstinence from ejaculation showed a significant increase in T2W and ADC values in the peripheral zone (adjacent to seminal vesicles) and resulted in larger seminal vesicle volumes and lower rates of nondiagnostic evaluation and therefore might improve evaluation of SVI^{69, 70}.

Figure 7.

Multiparametric MRI of a 51-year-old man with a serum prostate specific antigen level of 95.9 ng/mL. PI-RADS 5 lesion (asterisk) with frank capsular breach (arrows). Targeted biopsy of the lesion revealed Gleason score 10 (5+5) prostate adenocarcinoma. Seminal vesicles are invaded bilaterally (arrowheads). Targeted biopsy from the left and right seminal vesicles revealed Gleason score 7 (4+3) and Gleason score 8 (4+4) prostate adenocarcinoma, respectively. Axial T2-weighted imaging of the prostate (A), apparent diffusion coefficient map of the prostate (B), high b-value (b=1500 s/mm²) diffusion-weighted imaging of the prostate (C), dynamic contrast enhanced imaging of the prostate (D), axial T2-weighted imaging of the seminal vesicles (E), apparent diffusion coefficient map of the seminal vesicles (F), high b-value (b=1500 s/mm²) diffusion-weighted imaging of the seminal vesicles (G), dynamic contrast enhanced imaging of the seminal vesicles (H), sagittal T2-weighted imaging of the seminal vesicles (I), and coronal T2-weighted imaging of the seminal vesicles (J).

Similar to EPE evaluation, the risk of SVI was traditionally assessed using clinical nomograms, such as the Partin table⁴³ and Kattan nomogram⁷¹. These clinical monograms have moderate accuracy in detecting SVI^{31, 72}. Further studies have shown that MRI alone may perform similarly or superior to clinical assessments, and MRI plus clinical models can achieve the highest diagnostic accuracy for SVI detection^{45, 73, 74}. A recent meta-analysis of the literature with 34 studies has shown that, although mpMRI provides a low sensitivity (58%), the specificity is quite reliable for SVI detection (96%)⁵¹. Additionally, in a study conducted by Kim et al.⁷⁵, among a total of 1403 patients who underwent preoperative mpMRI, it was concluded that the sensitivity and specificity for corresponding pathological SVI was 45% and 95%, respectively. Furthermore, one study showed that there is no huge variance in SVI interpretation for radiologists with different reading experiences⁷⁶. SVI interpretation between a new radiologist and a senior radiologist showed a sensitivity of 18% vs. 27%, a specificity of 97% vs. 94%, a positive predictive value of 40% vs. 32%, a negative predictive value of 92% vs. 93%, and an overall accuracy of 90% vs. 88%, respectively. Overall, current data suggest that mpMRI provides a low sensitivity but a high specificity for detecting SVI. Nonetheless, studies have suggested that mpMRI may be comparable to, or even superior to, traditional clinical monograms, rendering it a valuable asset for ruling in SVI, staging prostate cancer, and guiding treatment decisions.

Biparametric MRI (bpMRI) is becoming more popular as the current trend is moving towards prostate MRI without contrast agent⁷⁷. This also raised questions regarding whether there is still a need for contrast agent for detecting SVI. In a retrospective study by Soylu et al.⁷⁸, two radiologists estimated the likelihood of SVI in three image-viewing settings: T2WI alone, T2WI plus DWI, and T2WI plus DWI and DCE sequences. When reviewing T2WI alone, both radiologists achieved high specificity (93% and 94%) and high negative predictive value (94% and 95%) but only moderate sensitivity (52% and 59%) and positive predictive value (50% and 52%). The addition of DWI significantly improved specificity (97% and 98%) and positive predictive value (70% and 79%) for both radiologists. However, adding DCE showed no improvement, suggesting that DCE may not provide incremental value for the diagnosis of SVI. Christophe et al. ⁷⁹ also demonstrated similar findings with AUC of 0.73 (95% CI: 0.66-0.83) for mpMRI and 0.76 (95% CI: 0.68-0.85) for bpMRI. Another more recent retrospective study showed conflicting results, however. Caglic et al.⁸⁰ found that mpMRI with DCE was better at detecting SVI and improved the inter-reader agreement compared to bpMRI without contrast. The AUC for mpMRI was 0.91 (95% CI: 0.84-0.96) and for bpMRI was 0.86 (95% CI: 0.78-0.92, p = 0.02), while the inter-reader agreement for prediction of SVI was 0.75 (95% CI: 0.62-0.88) for mpMRI and 0.69 (95% CI: 0.57-0.81) for bpMRI. The discrepancy between studies may be due to the difference in patient selection and image acquisition techniques. Caglic et al. examined biopsy naïve patients only whereas Soylu et al. conducted their study in surgical patients, who likely have more advanced diseases and larger lesions on MRI. Soylu et al. also used endorectal coils while Caglic et al. did not. Both studies were retrospective with limited sample size. Larger prospective studies are needed to further evaluate how DCE sequence impact the detection of SVI.

Similar to categorical EPE grading systems, Jung et al.⁸¹ proposed a 6-class grading system for detection of SVI using T2WI morphologic features from exams performed on 1.5 T scanner with endorectal coil. Class 0 indicates normal findings whereas class 5 indicates apparent mass lesion with destructive architecture. For class 4 and 5 combined, the sensitivity was 71% with specificity of 97%. This study was limited by its retrospective nature and sample size of 217 patients. With improvement of mpMRI and introduction of more advanced hardware, such as 3 T scanners, this system should be refined (such as incorporating other mpMRI pulse sequences) and validated in external cohorts. In conclusion, SVI is a crucial factor in prostate cancer staging and prognosis, as it signifies a higher disease stage and worse outcomes. While traditional clinical nomograms have moderate accuracy in detecting SVI, mpMRI has been shown to be able to achieve higher overall diagnostic performance, making it a valuable tool for ruling in SVI and guiding treatment decisions. Despite mpMRI's low sensitivity, its high specificity for detecting SVI establishes it as a dependable diagnostic tool. Nonetheless, a standardized SVI grading system should be considered and developed to further improve the uniformity of SVI interpretations and performance of MRI in detecting SVI.

Evaluation of Positive Surgical Margin

Prostate margin involvement status is an important determinant of patient outcome after radical prostatectomy. Around 20-30% of cases are reported to show a PSM for prostate cancer after prostatectomy⁸². In fact, the presence of a positive margin is associated with a higher risk for biochemical recurrence and patients should be followed closely.t⁸³. While preventing PSM is one of the main objectives of prostatectomy, preserving the neurovascular bundles and membranous urethra is critical for preserving continence and potency. During a prostatectomy, PSMis defined as tumor cells located at the edge of the tissue resected. ⁸⁴. A study suggested that MRI lesions located at the apex are associated with a higher risk of PSM⁸⁵. This might be due to the lack of a well-defined capsule at the prostatic apex and due to the absence of a clear boundary between the apex and the urethral sphincter. Posterolateral section is another commonly involved site (Figure 8), and positive margin at this site may confer greater risk of biochemical recurrence when compared to those with negative surgical margins at this site⁸⁶.

Figure 8:

Multiparametric MRI of a 71-year-old man with a serum prostate specific antigen level of 6.5 ng/mL. The lesion (arrows) was 1.8 cm in the right apical peripheral zone located in the posterolateral section and assigned PI-RADS category 5. Radical prostatectomy showed Gleason score 8 (4+4) prostate adenocarcinoma with extra-prostatic extension. Positive surgical margins were also noted. T2-weighted imaging (A), apparent diffusion coefficient map (B), and high b-value (b=1500 s/mm²) diffusion-weighted imaging (C).

PSMs are often associated with the index and larger lesions⁸⁷. As mpMRI can identify such lesions, the information obtained from mpMRI can be utilized during surgery to target large lesions to decrease the incidence of PSM. Furthermore, preoperative prostate mpMRI can identify patients with risk factors associated with PSMs and help surgical planning. In a recent retrospective study by Quentin et al.⁸⁸, it was concluded that a large percentage of PSMs were located at the apical and posterior capsule. In this study, mpMRI visualized 80% of prostate cancer with PSM at the urethra and 100% at the bladder.The length of capsular contact was the best MRI predictor for PSM at the capsule. However, for tumors with PSMs at the apical urethra, the distance to the membranous urethra was the best MRI parameter. The highest accuracy was documented in cases in which the distance between the prostate cancer and membranous urethra was ≤ 3.5 mm, indicating a high risk for PSMs at the urethra, and for the length of capsular contact ≥ 22.5 mm, indicating a high risk for PSMs at the capsule. Similarly, Park et al.⁸⁹ developed a scoring system to estimate the risk of PSMs based on MRI features. The system assigned scores as follows: PI-RADS categories 1-2 received 0 points, categories 3-4 received 2 points, and category 5 received 3 points. Tumors located at the posterolateral side or apex were each assigned 1 point. Capsular contact length between 15-24 mm was given a score of 1, while lengths of ≥ 25 mm were scored 2. The cumulative score ranged from 0 to 7 points, with a higher score indicating an increased risk of PSM. The scoring system exhibited strong predictive performance for PSM in both the derivation (C statistics, 0.80) and validation (C statistics, 0.77) groups. By utilizing a cutoff score of ≥ 4 for predicting PSM, the sensitivity, specificity, positive predictive value, and negative predictive value were determined to be 63%, 72%, 51%, and 80%, respectively. It is important to note that, one single-institution randomized trial with 438 patients did not show a clear reduction of PSM rate by performing MRI before surgery⁹⁰. However, in this study, the observed frequency of PSM was roughly half of what was anticipated, which may have resulted in insufficient statistical power to establish a significant impact of MRI. Additionally, one meta-analysis suggested that while the decision-making process for determining the extent of resection during radical prostatectomy is significantly influenced by MRI, it is unclear whether preoperative MRI effectively reduces PSM rates⁹¹. Due to the conflicting results from prior studies, the impact of mpMRI on detecting PSM should be further explored.

In summary, PSM after radical prostatectomy remains as a significant concern as it is associated with an increased risk of early biochemical recurrence for which a close clinical follow-up is required. The utilization of preoperative mpMRI enables the assessment of both pelvic anatomy and tumor positioning, which can aid in mitigating unfavorable surgical outcomes like PSMs. The MRI features such as tumor-capsule contact length, PI-RADS category, and tumor location have been used to develop predictive scoring systems for PSM. However, the impact of mpMRI on reducing PSM rates is still unclear and warrants further investigation. Overall, the integration of mpMRI into clinical decision-making for prostate cancer management holds great value in improving clinical outcomes.

Conclusions

Traditional methods of prostate cancer staging, such as digital rectal examination, serum PSA testing, and clinical nomograms, have limitations in their ability to provide a comprehensive evaluation of the extent and aggressiveness of the disease. Due to the ability of MRI to provide high spatial resolution and soft tissue contrast, it has emerged as a promising tool for preoperative local T-staging of prostate cancer. The specificity of mpMRI in detecting EPE and SVI has been shown to be excellent, while the sensitivity remains modest. The use of various scoring systems may improve EPE detection and provide uniformity in the assessment across various clinical settings. Information from preoperative MRI can guide treatment planning, including surgical resection and radiation therapy. Despite its potential benefits, the use of MRI for prostate cancer staging also has some limitations. One of the main drawbacks is the high cost of the imaging technique, which may not be feasible for some patients or healthcare systems. Additionally, the interpretation of MRI is highly dependent on the expertise of the radiologist, and the technique requires specialized training and equipment. With the increased focus on MR image quality and the introduction of the Prostate Imaging Quality (PI-QUAL) score system, more studies are needed to examine the potential effects of image quality on local staging. Furthermore, artificial intelligence-based solutions can provide consistent and objective image evaluation. Most of the current research focuses on the identification of intraprostatic lesions, but with the development of better algorithms, local staging using artificial intelligence models could be a possibility in the near future. Overall, the use of MRI for local staging of prostate cancer represents a significant advance in the field of urologic oncology. As technology and imaging techniques continue to evolve, it is likely that MRI will play an increasingly important role in the diagnosis, staging, and management of prostate cancer. As such, it is important for clinicians and researchers to continue to explore the potential benefits and limitations of MRI in the local staging of prostate cancer, with the ultimate goal of improving patient care.

Key Points.

• Multiparametric MRI (mpMRI) has become the preferred imaging modality for local staging of prostate cancer and has been shown to provide additional information on the site and extent of disease compared to traditional clinical nomograms.

• The Prostate Imaging Reporting and Data System (PI-RADS) standardizes prostate mpMRI acquisition, interpretation, and reporting. It is used to characterize and assess focal intraprostatic lesions seen on mpMRI based on the probability of clinically significant cancer in treatment naïve man.

• While mpMRI assessment of extra-prostatic extension and seminal vesicle invasion demonstrates moderate sensitivity and positive predictive value, it exhibits high specificity and negative predictive value. Utilizing standardized systematic evaluation methods can further enhance the accuracy of evaluating both extra-prostatic extension and seminal vesicle invasion.

• Utilizing mpMRI for surgical planning can help mitigate unfavorable surgical outcomes by identifying large index lesions that are often associated with positive surgical margins.

Clinical Care Points.

• Extra-prostatic extension (EPE), seminal vesicle invasion (SVI), and positive surgical margin (PSM) indicate locally advanced disease and carry a worse clinical prognosis which can impact the choice of treatment.

• EPE can be visualized on multiparametric MRI (mpMRI) as bulging/irregularity of the capsule, broad curvilinear capsular contact length, or frank invasion of the surrounding structures.

• SVI can be identified on mpMRI as hypointense filling defects on T2WI and ADC maps, hyperintensity on high b value DWI, and early contrast enhancement on DCE within and/or along the seminal vesicle. Additional feature includes obliteration of the angle between the base of the prostate and the seminal vesicle.

• PSM is associated with large index lesions, especially for those located at the apex and posterolateral section of the prostate.

• The current treatment options for patients with locally advanced high-risk prostate cancer are radiotherapy with androgen deprivation therapy or radical prostatectomy. However, research is still ongoing to identify the optimal treatment approach for these patients.